How to achieve: development with DWM1001, production with DW1000

In order to be commercially viable, the project I’m working on needs to keep a tight rein on component cost.

At this stage, I’m pretty sure that architecturally, I can make do with a DW1000 IC (at about US$6 per). But in order to put together dev boards, I have to use something physically bigger - obviously, the DWM1001 module (US$14).

Working with the DWM1001 module, is it still possible for me to work at the IC level such that the working prototype can seamlessly transition to the DW1000 IC? If so, how do I go about doing this?

Hope this makes sense.
TIA
Josh

What production volumes are you looking at and where will it be produced? That will determine the most feasible and economic path forward to production.

Your question implies a very strong price sensitivity since working at the chip level requires parts beyond the DW1000 that the DWM1001 already has (like processor, crystals, regulators, PCB, etc). In order to beat the DWM1001 price for an all up operating device, you need to be looking at high volumes. The DWM1001 is not a complete device by itself, lacking battery, regulator, enclosure, etc, so building a purpose made device based on the DW1000 chip itself can lower some of those costs.

There are reasons to not use a DWM1001 in a final product, particularly one of high volume, such as size, fit, form, lacking features, etc, but price by itself is a much harder parameter on which to win.

It is definitely possible, I would even say easy in most cases.

If the DWM1001 contains the parts you want to use in your final design, namely the nRF52832 processor and the LSI2DH12 accelerometer (if you need one), then work with the DWM1001 to develop your system and then build your final device using the DWM1001 schematic as the starting point for your design. You could literally make your final product be identical operationally with the DWM1001 such that the same software loads on both and it operates the same even if the physical form is changed.

Typically, though, the DWM1001 doesn’t have everything you want (like, say, having an LED). Those can be added to a DWM1001 carrier board or hacked into an existing one to provide a development platform. That can then easily be converted to a final design for manufacture.

If, for cost reasons, you want to change the nRF52832 processor on the DWM1001 to something else, now it becomes a little more tricky, but still doable. There are plenty of low cost microcontrollers out there, and the final hardware design can use them, but you will have to port your software to the new processor chip. This usually involves changes to the peripheral drivers (SPI, GPIO, etc). At the volumes you imply, the cost of this port is paid back by the lower cost of the microcontroller in production.

If someone has a working DWM1001 based product and wants that converted to a custom made device ready for high volume production, that is generally an easy project to do. You can have your engineers do it or hire someone to provide you the design. The tricky bits won’t be the circuit design itself, but the enclosure, battery, other features you need. There is a lot of work to make a truly low cost high volume product and much of it is shopping for the lowest cost parts and suppliers.

In summary, don’t fear developing with the DWM1001. Whatever you get as an outcome can be converted into a high volume low cost design.

Mike Ciholas, President, Ciholas, Inc
3700 Bell Road, Newburgh, IN 47630 USA
mikec@ciholas.com
+1 812 962 9408

1 Like

Mike:

Thanks very much for this very informative response.

In short, the volumes won’t really be massively high, but the project is intended to be a (unique) consumer system and I have to make the effort to ensure its on-the-shelf price doesn’t spiral out of control to the point where stores won’t want to carry it.

I’m aware of the relative “critter comforts” offered by the DWM1001 and it’s an obvious no-brainer to use it for development. But my key motivator lies in scaling down the solution to fit the problem - which basically can be solved with knowing just distances. IOW, my line of thinking is “if I don’t need the extra locationing capabilities of the DWM1001, can I get by with just the IC instead and thereby save some materials costs?”

Josh

You will have to judge the product cost per unit versus the engineering cost to develop it. As a general rule, the lower the unit cost, the higher the engineering cost.

The DW1000 chip really has no locating ability intrinsically built into it. When you boil it down, all the DW1000 really does is precisely time transmissions and precisely time receptions against a local clock. All of the ability to turn those times into useful outputs, be it distance or location, come about from math performed in a processor outside the DW1000 chip.

Since the DW1000 cannot operate stand alone in any capacity, there is always a processor in the node somewhere to configure and operate it. If you design your product to use the DW1000 chip, you need to build a board with a processor of your choosing and program it to operate the DW1000 and do whatever math you require for the type of output you desire. In addition to the processor, you will need to also provide crystals, power regulation, passives, and antenna.

If you use the DWM1001, Decawave has already provided a processor for you, the Nordic nRF52832. The DWM1001 also includes the crystals, power regulation, passives, and antenna, so you don’t have to provide those.

In favor of a custom design is that you don’t need a carrier board for the DWM1001, your custom design can be all on one PCB which does save some costs over a DWM1001 plus carrier board design. Obviously, a custom design can be made to exactly fit your physical requirements and features you need.

With the information given, I’m unable to properly advise you on what the best path for your project is. We help many clients with these sorts of decisions. If you like, I’d be happy to have a call with you to discuss those particulars and provide that guidance. We can execute a non disclosure agreement prior to the call to allow more fruitful discussions which lead to more usable answers. Send me direct email to schedule if desired.

Mike Ciholas, President, Ciholas, Inc
3700 Bell Road, Newburgh, IN 47630 USA
mikec@ciholas.com
+1 812 962 9408